Ben,

On Tue, Sep 4, 2012 at 9:28 AM, Ben Goertzel <[email protected]> wrote:

> > It seems (to me) that the entire thrust of AGI is to DESIGN systems to be
> > able to mimic the EMERGENT properties of neural systems, rather than
> > "designing" systems to do other things, one of whose emergent properties
> > happens to be GI.
>
> Not really...
>
> Brains display a complex mix of "wired-in" versus emergent aspects, and so
> (hypothetically, if they worked as hoped/advertised when completed)
> would many AGI designs ...
>

No argument here, other than perhaps very slightly differing opinions as to
how much is wired-in vs. fundamental.

My point was that it is damn near impossible to even make a case for
designing in emergent properties.

*How would YOU do such a thing*, other than just trying zillions of
semi-random experiments? This would obviously require more experiments than
could be completed in any person's lifetime. It is my unsupported hope that
math will provide the answers while some of us are still alive, but just
like you can't make a good case for AGI working in your lifetime (lacking
really vast resources), I can't make a good case for developing the math in
my lifetime (lacking vastly more resources).

My one thin alternative thin hope is that if someone could develop a
measure for intelligence, with a logarithmic scale that went WAY down to
include a toaster, that a genetic algorithm might find the magic formula,
just as evolution worked in us. A GA approach has its own challenges, e.g.
it requires a computer with vastly more computing power than all of the
neurons in our brains, and it won't work unless clever engineers can
imagine ALL of the essential components needed for intelligence
(corresponding to the bits in the GA's "genome"). Neither of these two
requisites are anywhere near to being in place, so this approach cannot
even be started at this time.

Have you thought about a GA "supervisor" for OpenCog, to replace current
developers?!!! If the prospect of an OpenCog-descendant AGI is workable
(you already know my opinion), this might provide your only realistic shot
at it. As in the previous paragraph, you will first need the measure for
intelligence as part of this.

BTW my son Eddie is VERY interested in and enthusiastic about GA approaches
to researching intelligence. He thinks we are ALL a bit crazy to be
manually fiddling with whatever we are working on, rather than just turning
it over to a GA to do the fiddling for us. As much as I enjoy fiddling, I
must admit that he has some really strong arguments.

I have tried to draw Eddie into discussions here, but apparently we haven't
yet made a good enough case for our own sanity.  8-:D>

Steve



-------------------------------------------
AGI
Archives: https://www.listbox.com/member/archive/303/=now
RSS Feed: https://www.listbox.com/member/archive/rss/303/21088071-c97d2393
Modify Your Subscription: 
https://www.listbox.com/member/?member_id=21088071&id_secret=21088071-2484a968
Powered by Listbox: http://www.listbox.com

Reply via email to